Biomedical subjects

James X Song

Publications and source records attributed to James X Song.

6 recordsLinked to original sources

PubMed2005

Diagnostic criteria for postconcussional syndrome after mild to moderate traumatic brain injury.

This study evaluated the prevalence and specificity of diagnostic criteria for postconcussional syndrome (PCS) in 178 adults with mild to moderate traumatic brain injury (TBI) and 104 with extracranial trauma. Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) and International Classification of Diseases (ICD-10) criteria for PCS were evaluated 3 months after injury. The results showed that prevalence of PCS was higher using ICD-10 (64%) than DSM-IV criteria (11%). Specificity to TBI was limited because PCS criteria were often fulfilled by patients with extracranial trauma. The authors conclude that further refinement of the DSM-IV and ICD-10 criteria for PCS is needed before these criteria are routinely employed.

Adult↗

PubMed2004

Limited agreement between criteria-based diagnoses of postconcussional syndrome.

The objectives of this study were to compare diagnoses of postconcussional syndrome between the International Classification of Diseases, 10th revision (ICD-10) and Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV). The patient sample was comprised of 178 adults with mild-moderate traumatic brain injury (TBI). The study design was inception cohort, and the main outcome measure was a structured interview 3 months after injury. The results were that, despite concordance of DSM-IV and ICD-10 symptom criteria (kappa=0.73), agreement between overall DSM-IV and ICD-10 diagnoses was slight (kappa=0.13) because fewer patients met the DSM-IV cognitive deficit and clinical significance criteria. Agreement between DSM-IV postconcussional disorder and ICD-10 postconcussional syndrome appears limited by different prevalences and thresholds.

Adult↗

PubMed2003

An evaluation of methods for the stratified analysis of clustered binary data in community intervention trials.

A simulation study is conducted in a community intervention setting. Several methods of stratified analysis of clustered binary data are compared in terms of empirical significance and empirical power levels. They are the Mantel-Haenszel test statistic (chi(2) (MH)), the adjusted Mantel-Haenszel test statistic of Donald-Donner (chi(2) (DD)), Rao-Scott (chi(2) (RSN) and chi(2) (RSP)), and Zhang-Boos (chi(2) (ZBN) and chi(2) (ZBP)), Wald (chi(2) (W)), robust Wald (chi(2) (RW)), score (chi(2) (S)), robust score (chi(2) (RS)), and the test statistic based on generalized linear mixed model (GLMM) (chi(2) (GLMM)). When rho not equal 0, chi(2) (MH) has inflated type I error, and it should not be used when observations are correlated. The results also warn of the use of chi(2) (RSN) and chi(2) (RW) due to their poor performance in terms of empirical significance level. chi(2) (ZBP) and chi(2) (GLMM) have better empirical significance levels as compared to other statistics; however, chi(2) (ZBP) tends to have lower empirical powers than other statistics when the number of clusters (N) is less than 24. chi(2) (RSP) provides the highest empirical powers when rho > or = 0.1 and N < or = 12. When rho < or = 0.01, we recommend the use of chi(2) (RS) and chi(2) (GLMM) since they have better overall performance in terms of empirical significance levels and empirical power levels.

Adolescent↗

PubMed2003

Sample size for K 2x2 tables in equivalence studies using Cochran's statistic.

This paper presents a new sample size formula for Cochran's test that uses additional information on stratum-specific success rates and requires fewer subjects for an equivalence study. Equivalence studies are common in clinical trials, where unlike superiority studies, the goal is to show whether a new drug therapy is as effective as a standard one. Stratification is typically used to adjust for differences among individual clinical trial centers with different success rates. The sample size is derived for a clinical trial design where two independent binomial proportions are compared within each stratum. Implementation of the sample size formula is described when the number of centers is large and the success rates of each individual center are not known exactly. The effect of variability of the success rates on the power of Cochran's test is shown through simulation. The variability of the success rates is measured by the intracluster correlation coefficient, which can be estimated by the ANOVA estimator of Donald and Donner. The simulation results show that the new sample size formula requires fewer subjects than sample size methods, which ignore stratification. The new method provides greater savings as the variability of success rates among centers increases.

Analysis of Variance↗

PubMed2003

Subject ordered pointing task performance following severe traumatic brain injury in adults.

The utility of a non-verbal, untimed subject ordered pointing task for identifying memory deficit in adult patients with TBI was tested. Using a cross-sectional design, the working memory performance of 70 adults with severe traumatic brain injury (TBI) and 45 uninjured adults was investigated on a computerized, self-paced, non-verbal subject ordered pointing task. Persons with severe TBI were impaired on measures of working memory relative to the control subjects. In addition, the task appeared to be sensitive to severity of injury as measured by the Glasgow Coma Scale, even within a truncated range of severity (GCS scores 3-8). It was concluded that the subject ordered pointing task is useful in identifying memory deficits in persons with brain injury.

Adult↗

PubMedDate not supplied

Inference methods for saturated models in longitudinal clinical trials with incomplete binary data.

In the longitudinal studies with binary response, it is often of interest to estimate the percentage of positive responses at each time point and the percentage of having at least one positive response by each time point. When missing data exist, the conventional method based on observed percentages could result in erroneous estimates. This study demonstrates two methods of using expectation-maximization (EM) and data augmentation (DA) algorithms in the estimation of the marginal and cumulative probabilities for incomplete longitudinal binary response data. Both methods provide unbiased estimates when the missingness mechanism is missing at random (MAR) assumption. Sensitivity analyses have been performed for cases when the MAR assumption is in question.

Algorithms↗